Transferable Atomic Multipole Machine Learning Models for Small Organic Molecules.

  title={Transferable Atomic Multipole Machine Learning Models for Small Organic Molecules.},
  author={T. Bereau and Denis Andrienko and O. Anatole von Lilienfeld},
  journal={Journal of chemical theory and computation},
  volume={11 7},
Accurate representation of the molecular electrostatic potential, which is often expanded in distributed multipole moments, is crucial for an efficient evaluation of intermolecular interactions. Here we introduce a machine learning model for multipole coefficients of atom types H, C, O, N, S, F, and Cl in any molecular conformation. The model is trained on quantum-chemical results for atoms in varying chemical environments drawn from thousands of organic molecules. Multipoles in systems with… 

Figures and Tables from this paper

Computer-aided drug design, quantum-mechanical methods for biological problems.
Developing Potential Energy Surfaces for Graphene-based 2D-3D Interfaces from Modified High Dimensional Neural Networks for Applications in Energy Storage
  • Vidushi Sharma, D. Datta
  • Computer Science, Materials Science
    Journal of Electrochemical Energy Conversion and Storage
  • 2022
Improved accuracy in ML-based modeling approach promises cost-effective means of designing interfaces in heterostructure energy storage systems with higher cycle life and stability.
Atomistic Simulations for Reactions and Vibrational Spectroscopy in the Era of Machine Learning─Quo Vadis?
  • M. Meuwly
  • Physics
    The journal of physical chemistry. B
  • 2022
Atomistic simulations using accurate energy functions can provide molecular-level insight into functional motions of molecules in the gas and in the condensed phase. This Perspective delineates the
Deep Potentials for Materials Science
To fill the gap between accurate (and expensive) ab initio calculations and efficient atomistic simulations based on empirical interatomic potentials, a new class of descriptions of atomic
A deep potential model with long-range electrostatic interactions.
Machine learning models for the potential energy of multi-atomic systems, such as the deep potential (DP) model, make molecular simulations with the accuracy of quantum mechanical density functional
Learning Atomic Multipoles: Prediction of the Electrostatic Potential with Equivariant Graph Neural Networks.
The accurate description of electrostatic interactions remains a challenging problem for classical potential-energy functions. The commonly used fixed partial-charge approximation fails to reproduce
Machine learning potentials for extended systems: a perspective
The present status of these new types of models for extended systems, which are increasingly used for materials modelling, are summarized and remaining challenges and limitations of current approaches are discussed.
Artificial Intelligence Designer for Highly-Efficient Organic Photovoltaic Materials.
This work describes an automatic design framework based on an in-house designed La FREMD Fingerprint and machine learning algorithms for highly efficient OPV donor molecules that demonstrates the ability to design new materials based on the substructure-property relationship built by ML.
Machine-Learning Interatomic Potentials for Materials Science
Large-scale atomistic computer simulations of materials rely on interatomic potentials providing computationally efficient predictions of energy and Newtonian forces. Traditional potentials have


Toward transferable interatomic van der Waals interactions without electrons: the role of multipole electrostatics and many-body dispersion.
The Voronoi tesselation approach to polarizabilities of atoms in molecules without electron density achieves an accuracy well within conventional molecular force fields while exhibiting a simple parametrization protocol.
Theory of intermolecular forces
Comparative Study of Selected Wave Function and Density Functional Methods for Noncovalent Interaction Energy Calculations Using the Extended S22 Data Set.
None of the DFT methods fulfilled the required statistical criteria proposed in this work, they cannot be generally recommended for large-scale calculations, and the best performing WFT methods were found to be the SCS-CCSD and MP2.5.
Benchmark database of accurate (MP2 and CCSD(T) complete basis set limit) interaction energies of small model complexes, DNA base pairs, and amino acid pairs.
MP2 and CCSD(T) complete basis set (CBS) limit interaction energies and geometries for more than 100 DNA base pairs, amino acid pairs and model complexes are for the first time presented together.
Representative Amino Acid Side Chain Interactions in Proteins. A Comparison of Highly Accurate Correlated ab Initio Quantum Chemical and Empirical Potential Procedures.
High degree of agreement was found between the different methods, even though the range of binding energies obtained was extremely large, and among the less computationally time-consuming methods, the DFT-D method as well as parm03 force field provided consistently good results when compared to the reference values.
Quantum Mechanical Calculations for Benzene Dimer Energies: Present Problems and Future Challenges.
Factors influencing quantum mechanical calculations of nonbonded interactions between organic molecules are still imperfectly understood. Much effort has gone into efforts to calculate the structures
A density‐functional study of the intermolecular interactions of benzene
We have tested the performance of three frequently used density functionals (LDA, LDA+B, and LDA+B+LYP) in a study of the intermolecular interactions of benzene. Molecular geometries are
Polarisable multipolar electrostatics from the machine learning method Kriging: an application to alanine
We present a polarisable multipolar interatomic electrostatic potential energy function for force fields and describe its application to the pilot molecule MeNH-Ala-COMe (AlaD). The total